2023
DOI: 10.1016/j.ecoinf.2022.101931
|View full text |Cite
|
Sign up to set email alerts
|

Sweet potato leaf detection in a natural scene based on faster R-CNN with a visual attention mechanism and DIoU-NMS

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

2
7
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
5
2
1
1

Relationship

0
9

Authors

Journals

citations
Cited by 37 publications
(11 citation statements)
references
References 25 publications
2
7
0
Order By: Relevance
“…As discussed by Li et al [29], the inconsistent sizes of rectangular labeling boxes and labeling di culties could also decrease model performance. The presence of weeds, young leaves, and complex leaf scenes resulting in appearance distortion could lead to detection failures and increased FP ratings [30]. However, under the eld conditions examined in this study, our model was not affected by plants other than beans, as the elds exhibited good agronomic practices to control weeds (See Fig.…”
Section: Plant Detection Accuracymentioning
confidence: 94%
See 1 more Smart Citation
“…As discussed by Li et al [29], the inconsistent sizes of rectangular labeling boxes and labeling di culties could also decrease model performance. The presence of weeds, young leaves, and complex leaf scenes resulting in appearance distortion could lead to detection failures and increased FP ratings [30]. However, under the eld conditions examined in this study, our model was not affected by plants other than beans, as the elds exhibited good agronomic practices to control weeds (See Fig.…”
Section: Plant Detection Accuracymentioning
confidence: 94%
“…Deep convolutional neural network models also have been widely applied as a reliable analytical method to detection and counting of target structure in several crops [28][29][30][31]. In particular, Faster R-CNN has shown promise as an effective method for detecting plant structures and/or counting individual plants using drone images [32][33][34][35].…”
Section: Introductionmentioning
confidence: 99%
“…Object detection techniques can be broadly categorized into traditional algorithms and deep learning-based approaches [4]. Traditional object detection methods [5] [6] were conventionally utilized for equipment identification and anomaly detection in substations prior to the advent of deep learning-based approaches.…”
Section: Related Workmentioning
confidence: 99%
“…This area has attracted extensive research from scholars who have contributed many innovative and impactful methods. As described in relevant literature, these modeling approaches accurately identify crops and weeds, thereby enhancing crop yields and reducing the use of chemical herbicides, consequently mitigating adverse ecological impacts [3][4][5] . However, despite the wealth of research, significant challenges in actual agricultural environments remain inadequately addressed: these models perform suboptimally when processing real-world data containing noise.…”
Section: Introductionmentioning
confidence: 99%